Re-processing of Shallow and Deep Crustal Reflection Seismic Data along BABEL Line 7, Central Sweden

(1)

Examensarbete vid Institutionen för geovetenskaper ISSN 1650-6553 Nr 241

Re-processing of Shallow and Deep Crustal Reflection Seismic Data along BABEL Line 7, Central Sweden

Hanieh Shahrokhi

Uppsala universitet, Institutionen för geovetenskaper Examensarbete E, 30hp i Geofysik

ISSN 1650-6553 Nr 241

Tryckt hos Institutionen för geovetenskaper, Geotryckeriet, Uppsala universitet, Uppsala, 2012

The BABEL project (Baltic And Bothnian Echoes from the Lithosphere) was a collaboration among British, Danish, Finnish, German and Swedish geoscientists to acquire deep-crustal reflection and wide-angle refraction data in the Baltic Shield and Gulf of Bothnia. In 1989, the collection of 2,268 km of deep marine reflection seismic data was carried out. BABEL line 7, one of several BABEL profiles, is the focus of this study and runs north of the Åland islands, in an E-W direction in the Bothnian Sea, east of the city of Gävle. The previous seismic image of the BABEL line 7 displays a considerable change in the reflectivity pattern from a weak reflective lower crust in the west to a more highly reflective lower crust in the east, interpreted to be due to a transition from a stiff crust to a plastic crust from the west to the east.

The seismic results were presented by the BABEL Working Group (1993) which focused on imaging and interpreting deep crustal structures as well as assessing the seismic velocities within the crust, the depth and nature of the Moho discontinuity and the seismic reflectivity texture in the crustal geological structures. Early Proterozoic plate tectonics in the Baltic Shield was also suggested from the reflection seismic data.

The BABEL line 7 reflection data were collected with a profile length of 174 km, a group of 48 air guns towed at 7.5 m depth, and 3000 m long streamer, 60 hydrophones spaced at 50 m intervals towed at 15 m depth. Seismic data were acquired for a 25 s record length using a 4 ms sampling interval and a 75 m shot interval. Seismic data are characterized by strong source-generated noise at shallow travel times and strong but randomly distributed spurious spikes at later arrival times.

In this thesis, the seismic data along BABEL line 7 were recovered and re-processed.

Modern processing techniques that were not available previously, were used. A special emphasis on the shallow parts of the seismic data was given and resulted in revealing reflections as shallow as 300 ms. Some of these reflections seem to be a continuation of the deeper ones and now appear to come to the surface which can now improve the correlation with the surface geology. Two major apparently moderately dipping shear zones are now interpreted to reach to the surface in the re-processed data in comparison with the previous work.

The deep reflections are also enhanced together with the improvement in the shallow parts which provide further insights about the nature of the Moho and its geometry along BABEL line 7. The re-processed seismic image demonstrates the potential in improving shallow and deep crustal structures along the BABEL offshore seismic data.

(2)

Examensarbete vid Institutionen för geovetenskaper ISSN 1650-6553 Nr 24

Re-processing of Shallow and Deep Crustal Reflection Seismic Data Dlong

BABEL Line 7, Central Sweden

Hanieh Shahrokhi

Supervisor: Alireza Malehmir

(3)

Tryckt hos Institutionen för geovetenskaper, Geotryckeriet, Uppsala Universitet, Uppsala 2012.

(4)

1

^Abstract

The BABEL project (Baltic And Bothnian Echoes from the Lithosphere) was a collaboration among British, Danish, Finnish, German and Swedish geoscientists to acquire deep-crustal reflection and wide-angle refraction data in the Baltic Shield and Gulf of Bothnia. In 1989, the collection of 2,268 km of deep marine reflection seismic data was carried out. BABEL line 7, one of several BABEL profiles, is the focus of this study and runs north of the Åland islands, in an E-W direction in the Bothnian Sea, east of the city of Gävle. The previous seismic image of the BABEL line 7 displays a considerable change in the reflectivity pattern from a weak reflective lower crust in the west to a more highly reflective lower crust in the east, interpreted to be due to a transition from a stiff crust to a plastic crust from the west to the east.

The seismic results were presented by the BABEL Working Group (1993) which focused on imaging and interpreting deep crustal structures as well as assessing the seismic velocities within the crust, the depth and nature of the Moho discontinuity and the seismic reflectivity texture in the crustal geological structures. Early Proterozoic plate tectonics in the Baltic Shield was also suggested from the reflection seismic data.

The BABEL line 7 reflection data were collected with a profile length of 174 km, a group of 48 air guns towed at 7.5 m depth, and 3000 m long streamer, 60 hydrophones spaced at 50 m intervals towed at 15 m depth. Seismic data were acquired for a 25 s record length using a 4 ms sampling interval and a 75 m shot interval. Seismic data are characterized by strong source-generated noise at shallow travel times and strong but randomly distributed spurious spikes at later arrival times.

In this thesis, the seismic data along BABEL line 7 were recovered and re-processed. Modern processing techniques that were not available previously, were used. A special emphasis on the shallow parts of the seismic data was given and resulted in revealing reflections as shallow as 300 ms. Some of these reflections seem to be a continuation of the deeper ones and now appear to come to the surface which can now improve the correlation with the surface geology. Two major apparently moderately dipping shear zones are now interpreted to reach to the surface in the re-processed data in comparison with the previous work.

The deep reflections are also enhanced together with the improvement in the shallow parts which provide further insights about the nature of the Moho and its geometry along BABEL

(5)

2

line 7. The re-processed seismic image demonstrates the potential in improving shallow and deep crustal structures along the BABEL offshore seismic data.

(6)

3

Acknowledgments

This thesis would not have been possible without the guidance and the help of many people who contributed and extended their valuable assistance in the completion of this study.

In the first place, I would like to express my gratitude to my supervisor, Dr. Alireza Malehmir who was abundantly helpful and offered invaluable assistance, support and guidance. Deepest gratitude is also due to Prof. Christopher Juhlin who without his knowledge and experience this study would not have been successful.

I gratefully acknowledge Daniel Sopher, Dr. Håkan Sjöström and Dr. Karin Högdahl for their comments, contributions, which made them a backbone of this study and so to this thesis.

I would like to thank to all my friends, especially: Pouya Ahmadi, Taher Mazloomian, Magnus Andersson, Faramarz Nilforoushan, Amir Abdi, Siddique Akhtar, Saeid Cheraghi, David Kellogg, Monika Ivandic and Saba Joodaki.

I wish also to acknowledge Prof. Annakaisa Korja and Prof. Raimo Lahtinen who kindly let me to use their figures in my thesis.

GLOBE Claritas^TM under license from the Institute of Geological and Nuclear Sciences Limited, Lower Hutt, New Zealand was used to process the seismic data.

Last but not least: I would like to express my love and gratitude to my beloved family: my father, mother and sister; for their understanding and endless love, through the duration of my studies.

(7)

4

List of figures

Figure 1.1. Geological map of the Fennoscandia ... 11

Figure 1.2. An example formation tectonic scenario of the Central Svecofennides ... 13

Figure 1.3. Bouguer anomaly map of the southern part of the Gulf of Bothnia Sea ... 15

Figure 2.1. Shot gathers, A) raw shot and B) spherical divergence ... 23

Figure 2.2. Shot gathers, A) deconvolution and B) band pass filter ... 24

Figure 2.3. Sketch illustrating the NMO geometry ... 26

Figure 2.4. Common mid point and common depth point concept ... 27

Figure 2.5. Stacked section before applying any poststack processing sequences ... 28

Figure 2.6. Stacked section showing the effect of poststack deconvolution ... 29

Figure 2.7. Stacked section shows the effect of FX-deconvolution ... 30

Figure 2.8. Stacked section shows the effect of scaling ... 31

Figure 2.9. Stacked section shows the effect of FK-mute ... 32

Figure 2.10. Stolt migrated seismic section ... 33

Figure 3.1. Stacked sections, A) previous result and B) re-processed result ... 36

Figure 3.2. A portion of the stacked section, A) previous and B) re-processed ... 37

Figure 3.3. A portion of the stacked section, A) previous and B) re-processed ... 38

Figure 4.1. Interpretation of the migrated section ... 40

Figure 4.2. A clear reflection from the mantle ... 41

(10)

7

List of tables

Table 2.1. Main acquisition parameters along BABEL line 7 ... 19 Table 2.2. Principal re-processing sequences with main parameters ... 20

(11)

8

List of abbreviations

CMP Common Mid-Point

CDP Common Dip Point

NMO Normal Move Out

DMO Dip Move Out

AGC Automatic Gain Control

TWT Two Way Time

SEG Society of Exploration Geophysicist

TREMOVE Totally REMOVE

SPHDIV Spherical Divergence

AVO Amplitude Versus Offset

FDFILT Frequency-Domain Filter

TVFILT Time-domain time-Varying Filter

FFT Fast Fourier Transform

PSDECON Post-Stack Wiener Deconvolution

FXDECON FX-domain complex Wiener Deconvolution

PS Phase – Shift

FD Finite Difference

BABEL Baltic And Bothnian Echoes from the Lithosphere

FIRE FInnish Reflection Experiment

SRME Surface Related Multiple Elimination

(12)

9

Chapter1 Introduction

The BABEL project (Baltic and Bothnian Echoes from the Lithosphere) was a collaboration among British, Danish, Finnish, German and Swedish geoscientists. At the beginning of 1980's, Dr. Drum Matthews along with several other scientists who were involved in BIRPS (British Institutions Reflection Profiling Syndicate), were trying to understand the reason for a very reflective lower crust in the western part of the Europe which were significantly different in the platforms and old shields e.g. in the United States according to COCORP land profiles results (Meissner et al., 1991).

The main problem was that there had not been any marine reflection seismic surveys on old shields and cratons and, thus, it was not clear if the reason for the contrast was due to the geological structure of the shield or due to the technical problems in the acquisition and processing. European geoscientists realized that they needed to execute a high resolution seismic reflection study. In January 1989, the final programs which were provided by scientists from five different nationalities were ready (Meissner et al., 1991).

Two main advantages of the BABEL project were the important geology of the Baltic Shield which allowed the group to use strong marine reflection acquisition, and the position of the Bothnian Bay which made the acquisition of both reflection and refraction data possible. This also provided the chance to employ onshore seismic refraction stations close to the reflection lines to record the data even in long offsets for obtaining velocity structures of the Baltic Shield (Meissner et al., 1991).

In September and October 1989, a total length of 2268 km of seismic data was acquired.

A 120 litre airgun array which was towed at 7.5 m depth and 3000 m streamer containing 60 groups of 64 hydrophones towed at 15 m depth were used. Sixty four multicomponent land stations recorded refraction data.

The record length in the northern part of the bay was 25 s (TWT) with 75 m shot interval and 18 s (TWT) in the southern part with a 50 m shot point interval (BABEL Working Group, 1991). The BABEL Line 7, the focus of this study, is about 174 km long and is located in the Bothnian Sea in an E-W direction (Fig 1.1). It runs from the east of the

(13)

10

city of Gävle in Sweden and passes north of the Åland Islands and ends in the west of the city of Turku in Finland. The shot interval was 75 m with 25 s record length. The line 7 crosses line 1 at about 128 km distance, line C at about 70 km distance and line 6 at about 50 km distance (Fig 1.1).

(14)

11

Figure 1.1. Geological map of the Fennoscandia based on Koistinen et al. (2001).

Upper inset: Major Paleoproterozoic orogens of the Fennoscandia. Lower inset:

Hidden and exposed microcontinental nuclei and arcs older than 1.92 Ga in Fennoscandia (based on Lahtinen et al., 2005), BABEL lines added to the geological map (modified after Lahtinen et al., 2008). The data along line 7 is the focus of this study and is shown in red colour.

(15)

12

1.1. Objectives

BABEL line 7 was processed after the data acquisition in 1989 and it has been interpreted and discussed in several papers (e.g. Korja and Heikkinen, 2005; Korja et al., 2001; Korja et al., 1993; BABEL Working Group, 1991). Figure 1.2, shows an example interpretation from the seismic data combined with surface geological observations (Korja and Heikkinen, 2005). However, since then no published attempt to re-process these data has been reported. BABEL line 7 is the closest line to a recent onshore reflection seismic survey that was conducted near the Dannemora iron mine (Malehmir et al., 2011). Therefore, an attempt to recover and re-process the data along BABEL line 7 was envisioned to potentially correlate structures observed on these data with those already observed on the line 7 (see also Juhlin and Stephens, 2006).

The main objectives of this study are (1) to improve the shallow parts of the seismic data for correlation with surface geological observations which was previously thought to be difficult, and (2) at the same time improving the deeper parts of the seismic section especially the Moho image. The seismic data were re-processed using Globe Claritas and obviously took the advantages of modern processing technologies (e.g. DMO corrections and prestack and poststack multiple attenuations) which were not available at that time.

In this thesis, it is shown how modern processing technologies have allowed imaging reflections as shallow as 300 ms. These reflections can now be followed from the surface to depths of about 30 – 35 Km (10 s), an important outcome of this thesis.

After a review of main data acquisition and processing steps (Chapter 2), the re- processing result will be presented and compared with the previous result (Chapter 3), and the interpretation (Chapter 4) of the data will be briefly presented and at the end main findings will be summarized in conclusion (Chapter 5).

(16)

13

Figure 1.2. An example formation tectonic scenario of the Central Svecofennides (adapted from Korja and Heikinnen, 2005). In this scenario, the central part of the Baltic Shield (Svecofennian part) was formed as the result of collision of the several micro continents (A-E). These interpretations are based on the results from BABEL seismic profiles shown in (F).

(17)

14

1.2. Review of general geology and tectonic setting

It is generally believed that the Fennoscandian Shield was formed during several phases of extensions and collisions (Korja et al., 1993) (Fig 1.2). The Archean craton and the Svecofennian Island Arcs collided and created the Svecofennian Orogeny (1.9 – 1.76 Ga); the result was a thick crust (50 – 65 km), then the thickness of the earth´s crust was reduced to about 45 km by an extensional phase in a E-W direction which also resulted in the intrusion of the Rapakivi granitoid (Korja et al., 1993). During this period the intruded anorogenic Rapakivi granitoid was formed beneath the Bothnian Bay (Korja et al., 1993). The thickening and thinning of the crust took place in relatively restricted time periods and does not support the idea that crustal thickness grows as a function of time (Korja et al., 1993).

The evolution of the Svecofennian orogeny began about 2.1 Ga ago with the rifting of the Archean domain, then the Svecofennian Island Arcs collided with the Archean craton, and finally the Rapakivi granitoids were emplaced at about 1.5 Ga ago during an extensional phase (Korja et al., 1993).

A study of the Bouguer anomaly map ranging from -20 to -40 mgal (Fig 1.3) measured in the northern part of the Baltic Sea in 1997 and the BABEL reflection seismic data provided evidence for the Subjotnian (1.6 – 1.5 Ga) palaeorift model of the crustal structure in the Svecofennian domain (Korja et al., 2001). The southern Gulf of Bothnia is part of Mesoproterozoic rift-related area which is composed of an old Svecofennian mafic intrusion in the lower to middle crust (Korja et al., 2001).

This was followed by Subjotnian Rapakivi granitoids, which shows dehydration melting in the upper crust (Korja et al., 2001). Finally, it was covered by the Jotnian sedimentary

rocks (Korja et al., 2001).

(18)

15

Figure 1.3. Bouguer anomaly map of the southern part of the Gulf of Bothnia Sea.

Outcropping Rapakivi granite batholiths are outlined in red (Adapted from Korja et al., 2001). BABEL line 7 is the focus of this study.

(19)

16

The crust reformation occurred during the Mesoproterzoic Subjotnian; anorogenic Rapakivi granite and gabbro intruded and then the Subjotnian to the Jotnian extensional, thin-skinned sedimentary basin was formed (Korja et al., 2001). The Postjotnian diabase dykes and sills (1.1 – 1.2 Ga) reveal the latest magmatic event which have similar geochemistry to the continental flood basalts (Rämö, 1991). The low topography, the crustal thickness gradients and large bimodal magmatism provide the case for palaeorift in the northern Baltic Sea (Korja et al., 2001).

(20)

17

Chapter 2 Data acquisition and processing

In spring 1989, when supports for the BABEL data acquisition was prepared, a commercial contractor was chosen to provide a long streamer, the seismic receiver unit, the acoustic source and a powerful airgun array (Meissner et al., 1991). Based on BIRPS experience with deep marine seismic data acquisition, PRAKLA-SEISMOS from Hannover was selected to bring their vessel (M/V MINTROP) (Meissner et al., 1991).

The streamer which was chosen should be adapted in both, the low density water in the Gulf of Bothnia as well as to the more dense water area in the southwest of the Baltic Sea (Meissner et al., 1991). A strong tuned airgun array of 120 litre, consisting of 6 sub arrays towed at 7.5 m depth and a 3000 m streamer designed with 64 hydrophones in 60 groups towed at 15 m depth was used (Table 2.1) (Meissner et al., 1991).

Applying different methods of signal processing is necessary after data acquisition in order to produce a seismic image that can be interpreted. Data processing sequences depends on the geological scenario, acquisition parameters and target depth. Processing of marine seismic data is different from land data since noise characteristics are quite different.

The seismic data processing usually consists of two main steps: prestack processing and poststack processing, where the main aim is to improve the signal to noise ratio.

The noise definition depends on the data application, so any recorded energy that interferes with expected signal is considered as noise (Elboth et al., 2008).

The noise can be classified in three groups: background noise (e.g., wind noise, swell noise, nearby production noise and interference from nearby seismic data acquisitions), seismic source generated noise (e.g., direct waves, scattered waves and multiples) and instrument noise (Elboth et al., 2008). All these noises can be classified in two main groups: coherent and random (incoherent) noises (Elboth et al., 2008):

 Coherent noise: can be due to back scattering, multiples and near surface ghosting, there are different methods to attenuate them e.g., different types of demultiples, deconvolution in Tau-p domain, deconvolution (both prestack and poststack) and

(21)

18

model based methods like SRME (Surface Related Multiple Elimination) (Stork et al., 2006).

 Random noise: can be reduced with different methods like FX-deconvolution which depends on a lot of parameters that can help to first recognize them and then remove or reduce their effects (Larner et al., 1983).

Table 2.2 summarizes main re-processing steps applied to the data. Here, some of the key processing steps and their effects are briefly explained.

(22)

19

Table 2.1. Main acquisition parameters along BABEL line 7 Survey area and year BALTIC SEA, 1989

Project BABEL

Line No. BABEL 7

Survey date 20.9.1989 – 07.10.1989 recorded by S. V. MINITROP

Tape date 30.11.1989

Tape version UK00A-P1-1984

Client NERC (BIRPS)

Geophysical contractor PRAKLA – SEISMOS AG NO.89202351

Positioning contractor RACAL – DECCA

Positioning processing PRAKLA – SEISMOS AG

Positioning system DECCA MC / MAGNAVOX – TRANSIT

Shot point position SHOTPOINT (CENTER SOURCE)

Offset antenna to shot 175 Meters

Spheroid as surveyed A, 1/F HAYFORD INT. 1924 6378388.000 .006722670

Shooting Dir. 85 DEG

Record length 25 s

Low-cut 3.0 Hz (6 dB/oct)

High-cut 102.0 Hz (484 dB/oct)

Streamer length 3000 m

Shot interval 75 m

Group spacing 50.0 m

Distance antenna array 176 m

Offset in line 119 m

Array depth 7.5 m

Streamer depth 15 m

Array type F120/090/7.5

Source volume 120.0 liter

Sample rate 4 ms

Field gain 12 dB

Fold coverage 20

(23)

20

Table 2.2. Principal processing sequences with main parameters, BABE line 7, 2012

1. Read SEG-Y format file – 25 s 2. Apply regular marine geometry

3. Trace editing (noisy and dead traces were killed) 4. Trace balancing

5. Spherical divergence corrections:

- Time – Velocity: 0 ms - 1500 m/s, 5000 ms - 6500 m/s, 25000 ms - 8000 m/s 6. Deconvolution:

- Window 200 – 5000 ms, Application window 0 – 3500 ms:

Max filter length = 135, max gap length = 30 ms, pre-whitening 0.1;

- Window 5000 – 25000 ms, Application window 5000 – 25000 ms:

Max filter length = 135, max gap length = 35 ms, pre-whitening 0.1 7. Time variant band pass filtering:

- 0 - 2500 ms: 0 – 10 – 100 – 120 Hz - 3000 – 25000 ms: 0 – 10 – 50 – 60 Hz 8. Common depth point (CDP) sorting 9. Velocity analysis (iterative)

10. Normal move out (NMO) correction, 50 percent stretch mute 11. Residual statics

- Surface consistent mode, time window length 1000 – 15000 ms 12. Dip move out (DMO) correction

- Bin value: -3100:0/10, Bin size: 5, CDP spacing: 25 m 13. Velocity analysis

14. Band pass filtering:

- 0 – 25000 ms: 10 – 12 – 50 – 60 Hz 15. Stack (median)

16. Poststack deconvolution

17. FX-deconvolution (FX-domain complex Wiener deconvolution) 18. Scale

19. FK-mute

20. Migration (Stolt migration)

- Using smoothed NMO velocity (5700 – 8200 m/s)

(24)

21

2.1. Spherical divergence corrections

A shot point is usually a source of the spherical waves (Yilmaz, 1989). The earth body has two effects on waves which are propagated through it (Yilmaz, 1989);

I- the energy density decreases according to 1/r² (r is wave radius) and we know that wave amplitude decay depends on the square root of energy density so it decays proportionately to 1/r in a homogeneous medium (Yilmaz, 1989). However, in inhomogeneous mediums like a layered earth amplitude decay will not be linearly dependent (Yilmaz, 1989). On the other hand, the velocity usually increases with depth which causes more divergence and also fast amplitude attenuation (Yilmaz, 1989).

II- the initial frequency of the source changes according to the time variant as it spreads in the medium (Yilmaz, 1989). Because of attenuation in rocks, high frequencies are absorbed faster than low frequencies (Yilmaz, 1989). To correct the attenuated signals that are far from the source, spherical divergence corrections with the parameters in table 2.2 were applied. The main problem with this processer is that it increases the amplitude of noise as well (Yilmaz, 1989). Figure 2.1 shows the effects of both trace balancing and spherical divergence corrections to a raw shot gather.

2.2. Deconvolution

Deconvolution described here, is based on optimum Wiener filtering (Claritas dictionary). It enhances the temporal resolution of the seismic data by compressing the raw seismic wavelet (Claritas dictionary). It is usually used before stack, although it can be applied after stack, it is applied both prestack and poststack in this study.

Deconvolution has two main effects on seismic data; first, compression of the wavelet to make it spikier and second, elimination of multiple energy, which is a significant problem in marine seismic data (Claritas dictionary).

To apply a Wiener deconvolution we filter assume:

I. the seismic trace is considered to be the convolution of the source wavelet with the reflection coefficients series (Claritas dictionary),

(25)

22

II. the reflection coefficient parameters are a white noise section (Claritas dictionary), III. the earth wavelet response and the seismic wavelet are minimum phase (Claritas

dictionary).

The main disadvantage of applying deconvolution is that it may decrease the amplitude of real reflections and can alter their amplitude and the phase (Claritas dictionary).

Two different design windows (200 - 5000 ms and 5000 - 25000 ms) were applied with two different application windows (0 – 3500 ms and 5000 – 25000 ms), the maximum filter length was set to 135 ms and gap length to 30 ms for first design window and 35 ms for second one (Fig 2.2A). As evident from Figure 2.2A, more reflections are now observed (see Fig 2.1B) which suggests successful application of the deconvolution filter.

2.3. Band pass filtering

FDFILT processor used in this study is a zero phase and time variant filter that transforms each trace to frequency domain and according to band pass filter boundaries makes some parts of the zero spectrum and leaves some parts unchanged. It uses a cosine taper to make smooth frequency changes among those frequencies which are passed and the ones which are removed. An inverse FFT is then applied to take it back to the time domain (Claritas Dictionary).

A time variant band pass filter was designed from 0 ms to 2500 ms with F1=0 H z, F2=10 Hz, F3=100 Hz and F4=120 Hz boundaries and from 3000 ms to 25000 ms and F1=0 Hz, F2=10 Hz, F3=50 Hz and F4=60 Hz boundaries (Fig 2.2B).

(26)

23

Figure 2.1 A) A raw shot gather from the western side of the line. The arrow on the top shows refracted arrivals. B) Spherical divergence correction is applied to (A); two clear sets of reflections are observed between 5 – 6 s.

(27)

24

Figure 2.2 A) After applying deconvolution to the shot gather shown in Figure 2.1B more reflections are observed and the reflection at shallow depth is better revealed. B) Band pass filter removed some low frequency noises (swell noises) on the channels which are closer to the source from 3 s to the end the record and allowed improving the continuity of the reflections marked by the arrows.

(28)

25

2.4. Velocity analysis and normal move out (NMO) corrections

The purpose of the velocity analysis is to bring out a velocity model that can flatten the reflected events using the NMO correction, a hyperbolic reflection in order to stack constructively (Yilmaz, 2001). A good velocity analysis and NMO correction is needed:

1) to improve signal to noise ratio in the stacked data, 2) to obtain a better time to depth conversion, and 3) to apply migration successfully (Yilmaz, 2001).

Irregularities in the elevation of the near surface within CDP gathers can cause delay in travel times of reflections. Residual statics is a time independent processing parameter as its name shows that correct these irregularities (Yilmaz, 2001).

To process the data, after first performing velocity analysis and finding an optimum NMO velocity model, the residual static corrections with 8 iterations were calculated and the best result was applied to the data. A second velocity analysis was then done in order to improve the initial NMO velocity model which was followed by another iteration of residual statics. The best result from these iterations was applied to the data prior to the stacking. The residual statics did not result in significant improvement as may be expected with marine seismic data.

The normal move out corrections is a function of time, offset and velocity (Yilmaz, 2001). A non-zero offset depends on the travel time and distance between source and receiver (offset) respectively (Yilmaz, 2001). The NMO correction (Fig 2.3) is used to correct all non-zero offset to a zero offset travel time (Yilmaz, 2001). The NMO corrections with 50% stretch mute were applied. The stretch mute is used because the normal move out correction stretches the traces in a time variant way and as a result their frequencies shift toward the low end of the spectrum. Frequency deformation increases at far offsets and at shallow travel times (Yilmaz, 2001).

(29)

26

Figure 2.3. Sketch illustrating the NMO geometry of a horizontal reflector and the NMO correction (Modified from Yilmaz, 2001).

2.5. Dip move out (DMO) corrections

The terms of CDP gather and CMP gather are different (Fig 2.4), this is due to lateral velocity variation and the dip of the subsurface reflector. The common mid-point (CMP) are dip-independent and remain the same for all shot-receiver pairs, but the depth point (D) in the common depth point (CDP) gather is different for each shot-receiver pair (Levin, 1971).

Each trace is migrated to zero offset with DMO correction and as the result, DMO corrected data can be greatly improved (Deregowski, 1986).

A prestack partial migration DMO corrections, with offset values between -3100 m to 0 m sliced every 10 m was applied, then a velocity analysis was done on the DMO corrected data and then this was used for NMO correction prior to the stacking of the data.

(30)

27

Figure 2.4. Common mid point (CMP) and common depth point (CDP) concept. DMO correction is required to move point D’ to D for the dipping reflector (from Yilmaz, 2001).

(31)

28

2.6. Stack

The seismic reflection data acquisition method is usually designed to produce multi-fold coverage for each shot. Using CDP sort, shot-receiver gathers are sorted to CDP gathers (Yilmaz, 2001).

The stack adds all the same mid-point traces together and makes a single trace for each ensemble. To process BABEL line 7, median stacking method was used for stacking (Yilmaz, 2001). Theoretically, the signal amplitude to rms noise ratio is enhance by a factor of √N with the assumption that reflected signals are identical and unchangeable and noise ratios change from trace to trace (Sengbush, 1983), but practically it does not work exactly like this, so the result is somehow less than √N. Stacking attenuates coherent noise like multiples (Mayne, 1962) because the stacking velocities for reflected signals and coherent noise are different (Yilmaz, 2001).

Figure 2.5 shows the stacked section before applying any poststack processing methods.

Figure 2.5. Stacked section before applying any poststack processing sequences.

(32)

29

2.7. Poststack deconvolution

Poststack deconvolution is performed as a trace by trace filter. It can be applied as a spike or gapped Wiener deconvolution (Claritas dictionary). It acts differently from prestack deconvolution since the autocorrelation function of each trace is smoothed before being used by the deconvolution filter (Claritas dictionary). As the result, we have less spatial variation and hence in data with lower signal to noise ratio weak signals are not damaged (Claritas dictionary).

A time window from 1000 ms to 25000 ms and a 500 ms filter length with 30 ms gap length and 0.1 percent pre-whitening was applied to remove multiples remained in the stacked section after stacking (Fig 2.6).

Figure 2.6. Stacked section showing the effect of poststack deconvolution.

(33)

30

2.8. FX-deconvolution

To attenuate random noise, FX-deconvolution, which is a poststack filter, was applied. It transfers each trace to frequency space, then a Wiener deconvolution is applied in X direction for each frequency, and finally the filter transfers the trace to the TX plane.

The output of the FX-deconvolution coherent filter is less messy than other post-stack filters, but the main problem of the filter is that its performance is in only one dip at a time, so other dips will be attenuated, because of that it works in a window area which can be defined by vertical and horizontal (T, X) panels (Claritas dictionary).

Twenty ms filter length and 100 traces with 100 ms time – window length was applied.

Figure 2.7 shows the attenuation of random noise by applying FX-deconvolution to the stacked section shown in Figure 2.6.

Figure 2.7. Stacked section shows the effect of FX-deconvolution on attenuating random noise.

(34)

31

2.9. Scale

The scale processing parameter is used to bring the amplitude values to an observable range, since we usually lose them in SPHDIV processing sequence. Because spherical divergence is a nonlinear processor and it attenuates amplitudes in shallow parts of the seismic data in the prestack domain, the scale is used to preserve all relative amplitudes and we can enhance amplitudes, by finding a suitable scaling function (Fig 2.8). Here, I used a window from 0 ms to 4500 ms, -70 dB and from 4500 ms to the end -20 dB scaling has applied.

Figure 2.8. Stacked section shows the effect of scaling which enhance the amplitude and enhance the reflections that were attenuated during other processing sequences.

(35)

32

2.10. FK-mute

Some specific types of undesired energy in the collected data with the dip in (t, x) plane can be eliminated by their dips in the (f, k) domain (Yilmaz, 1989). These kinds of coherent linear noises are usually isolated from the reflected energies in (f, k) space, so we can remove coherent noise by their dips from collected data (Yilmaz, 1989). An FK mute can be used both before and after stack (Yilmaz, 1989).

In Figure 2.9, an FK-mute has been able to attenuate some of the specific steep dips of the coherent noise in the stacked section.

Figure 2.9. Stacked section shows the effect of FK-mute in removing the coherent noises that have specific dip from the data.

(36)

33

2.11. Migration

One of the main goals of the processing of seismic data is to find the relation between the seismic image and the geological and structural settings of the studying area (Yilmaz, 1989). The migration method is used in order to move the reflections from their apparent locations in the stacked section to their true locations in the migrated section.

Four different kinds of 2-D migrations methods: Stolt frequency-wavenumber (f-k) poststack migration, PS or phase-shift poststack migration, FD or finite difference poststack migration and Kirchhoff poststack migrations were tested. The Stolt migration produced a better seismic image compared with the others, therefore, it was chosen (see Fig 2.9). A smoothed velocity function from 5700 m/s to 8200 m/s between 0 s to 15 s was applied. A stretch factor of 0.6 which is typically used for an increasing velocity with depth was used.

Figure 2.10. Stolt migrated section along BABEL line 7 (whole 25 s (TWT)).

(37)

34

Chapter 3 Comparison with previous work

The seismic reflection image from the previous processing of the BABEL line 7 (Fig 3.1A) does not show any clear reflection at shallow depth (0 - 2 s). According to a report from the Geological Survey of Finland, which compares the Finnish Reflection Experiment (FIRE) with the BABEL profiles, the reflectivity character of the uppermost and lowermost crust along the FIRE profiles are quite different from that in the BABEL lines. In the BABEL lines presented previously, usually the upper crust to a depth of 6 – 10 km is transparent or shows weak reflectivity, whereas in the FIRE lines (with some exceptions), the same depth is highly reflective (FIRE Working Group, 2006). As mentioned in the second chapter in the Svecofennian domain (Korja et al., 2001), the last tectonic event in that area was an extension when the Rapakivi granitoids were formed at about 1.5 Ga ago (FIRE Working Group, 2006). However, the reflectivity contrast between the FIRE and the BABEL profiles in the uppermost crust might not be because of this extensional event and the most plausible reason could be due to the different data acquisition and processing methods. It has been discussed that the powerful direct pressure wave produced by the airgun masks the uppermost crustal reflections gives the BABEL acquisition parameters (e.g., Fig 3.1A) by a streamer of 3000 m length long and the effect of the pressure would mask the first 2 seconds of the data, corresponding to about 6 km depth (FIRE Working Group, 2006).

A highly reflective shallow crust was observed in a short (30 km) land data survey in northern Sweden close to the Bothnian Gulf (Juhlin et al., 2002). The reflectivity of the upper part of the BABEL lines can be enhanced by processing techniques (Flueh and Dickmann, 1992) as demonstrated here by re-processing of BABEL line 7 (Fig 3.1B).

Figures 3.2A and 3.3B show portions of the stacked sections obtained using the previous processing sequence compare with the same portion obtained in this re-processing work.

As evident from (Fig 3.2B) two sets of east dipping reflections are imaged which are absent in the previous processing result (BABEL Working Group, 1993). Careful analysis of the shot gathers from the western most part of the line suggests that these reflections can also be observed in these data (e.g., Fig 2.1) which further confirms that these reflections (R1 and R2) are real and not artefacts from the data processing.

(38)

35

The main reason to reveal these reflections in the re-processing work can be due to the following:

1. The high frequencies in the shallow parts of the data have been retained

2. Improved stacking velocities; there were previously picked without DMO corrections.

The new work can show that shallow reflections are present in the data, and were preserved during the DMO corrections and poststack processing steps.

It is, however, clear that the sea-bed reflections (e.g., S1 and S2 in Figs 3.2A and 3.3A) are not presented in the re-processed image. The sea-bed reflections were not the main focus of this study; therefore, their absence in the final image is not significantly a problem.

A shallow reflection similar to R1 and R2 in the western part of the line is also imaged in the eastern part of the line (see Fig 3.3B).

The re-processed seismic section shows higher frequency content with stronger reflections than the previous seismic section (Fig 3.1A). The Moho discontinuity is better imaged in the re-processed data and shows continuous character along the entire line when compared with the previous result (Fig 3.1B). It is easier to interpret the reflections in the re-processed image than the previous one which again demonstrates the potential of re-processing these offshore data.

(39)

36

Figure 3.1. Stacked sections along the BABEL line 7, A) previous result (BABEL Working Group,1993) and B) re-processed result. The close ups of the seismic section are shown in Figure 3.2 and 3.3. Note that R1 and R2 can now be followed to the surface in the re-processed seismic section.

(40)

37

Figure 3.2. A portion of the stacked section (see Figure 3.1), A) previous result (BABEL Working Group, 1993), B) re-processed result. Note reflections R1 and R2 which are successfully imaged in the re-processing work (see text for detailed description of the marked events).

(41)

38

Figure 3.3. A portion of the stacked section (see Fig 3.1), A) previous result (BABEL Working Group, 1993), B) re-processed result. Note reflection R3 which is successfully imaged in the re-processing work (see text for detailed description of the marked events).

(42)

39

Chapter4

Interpretation

One of the aims of re-processing BABEL line 7 is to improve the image of the geological structure of the area; the Moho, lower crust and upper crust. The seismic reflection and wide angle refraction results of BABEL project have given quite clear images of deep geological structure and Moho boundaries in Bothnian Bay and Baltic Sea but there has been some discussion about why the uppermost crust in BABEL data acquisition is almost transparent and why at 5 km depth in, there is no reflection compared to the FIRE profiles which are highly reflective (Lahtinen et al., 2009). A suggested reason for this difference is the data acquisition environment (Lahtinen et al., 2009). In the BABEL survey a powerful airgun source is used which masks the weak surface reflections (Lahtinen et al., 2009). In this re-processing study, some reflections as shallow as 300 ms are imaged; R1, R2 and R3 (Fig 4.1) which prove that the new data processing methods can reveal the uppermost crustal reflections which suggests that the problem of not being able to see the shallow reflections in the BABEL data is due to the previous processing methods rather than the data acquisition.

4.1. Geological interpretation

In the western side of the profile, the lower crust is less reflective compared to the eastern side of the line which is highly reflective and the reflector dip is southeast oriented (Fig 4.1) (Korja and Heikkinen, 1995).

In the west part of the profile, in the upper crust, there is a southeast listric reflection which flattens out in the lower to middle crustal boundary. The upper and middle crust are fairly reflective with low angle reflectivity dipping towards southeast (Fig 4.1) (Korja and Heikkinen, 2005).

(43)

40

In the middle of the line, the upper crust is weakly reflective; however it shows some horst-graben formations. The middle crust has a strongly reflective pattern, here the junction between some listric reflections and the deeper most part of the reflection is also highly reflective, which has both dipping and flatten reflections and in the boundary shows a concave reflection structure (Korja and Heikkinen, 2005). At about 125 km distance, a clear reflection from the mantle is observable even after migration (Fig 4.2).

The stacked image shows a dramatic change from the very weakly reflective lower crust in the west of the line to the strongly reflective lower crust in the east, interpreted as the variation from a stiff crust to more plastic crust. Although it is not clear how much of the structure is original and how much of it has been shaped during the ―mafic mantle-derived magmatism‖ and Aland rapakivi granites formation (Korja and Heikkinen, 2005).

Figure 4.1. Interpretation of the migrated section; R1 and R2 are two shallow reflections with a gentle angle towards southeast can be seen in uppermost crust (they can be the continuous of the lower reflections). D1 is a southeast listric reflection in the upper crust which flattens out in the lower to middle crust boundary. M1 is a reflection from the mantle which can be seen beneath Aland Islands. R3 is a shallow reflection with SE direction in the uppermost crust.

(44)

41

Figure 4.2. M1 is a clear reflection from the mantle at about 125 km distance and 48 km depth.

(45)

42

Chapter 5 Conclusion

In 1989, the BABEL seismic data were collected by a multinational collaboration to study the geological structures of the Baltic Sea and the Gulf of Bothnia. A huge amount of marine seismic data was acquired, as well as onshore data from 64 multicomponent land stations for refraction data analysis. The results from the pre- processed data reveal a good number of reflections from the lower crust, the middle crust and the Moho in contrast with lowermost crust that almost shows no reflection in the 2 first seconds. By re-processing the BABEL line 7 (the E-W profile in the southern Gulf of Bothnia), with modern techniques and new methods, several reflections as shallow as 300 ms are imaged. There are two main shallow reflections in the western side of the data that could be a continuation of deeper reflections. This enables correlation of the deep events with near-surface geological structures. A shallow reflection similar to this is also imaged in the eastern side. The main reason for revealing these shallow reflections might be stacking velocities that were chosen carefully allowing the shallow reflections to be preserved during DMO and poststack processing.

The re-processed seismic image shows improved reflections when compared with the previous work and the Moho discontinuity is better defined along whole profile.

Considering those shallow reflections and the improvement of the image in the lower crust and at the Moho discontinuity, interpretation of the data is now easier.

Further work is recommended;

1) to re-process crossing lines (lines 6, 1 and C) with BABEL 7 to understand orientation of main reflections as well as their correlation;

2) to re-process all the BABEL profiles using advanced processing methods, the BABEL data still have a great potential to be improved in the lower crust and the Moho; but the main focus can be on the uppermost crust that shows poor image;

3) to study the correlation between water depth in the Baltic Sea and Bothnian Bay with shallow reflections coherency.

(46)

43

References

Abramovitz, T., Berthelsen, A., Thybo, H., 1997. Proterozoic sutures and terranes in the southeastern Baltic Shield interpreted from BABEL seismic data. Tectonophysics, 270, 259–277.

BABEL Working Group, 1991. Evidence for early Protertozoic plate tectonics from seismic reflection profiles in the Baltic Shield. Nature, 348, 34-38.

BABEL Working Group, 1993a. Integrated seismic studies of the Baltic Shield using data in the Gulf of Bothnia Region. Geophys. J. Int. 112, 305-324.

Deregowski, S. M., 1986. What is DMO: First Break, 04, 7- 24.

Elboth, T., Qaisrani, H. & Hertweck, T., 2008. De-noising seismic data in the time-frequency domain. SEG Technical Program Expanded Abstracts, 27(1),2622-2626.

Flueh, E.R., & Dickmann, T. 1992. Technical aspects of wide-angle data collection and processing. In: The BABEL project. Commission of the European Communities, ed.

By R. Meissner et al., Directorate-General, Science, Research and Development, Brussels, Commission of the European Communities, pp. 123–130.

Gaal, G., Gorbatschev, R., 1987. An outline of the Precambrian evolution of the Baltic Shield.

Precambrian Res. 35, 15–52.

Geological Survey of Finland, 2006. Ilmo T. Kukkonen, Pekka Heikkinen, Elias Ekdahl, Sven-Erik Hjelt, Jukka Yliniemi, Erkki Jalkanen and FIRE Working Group.

Gohl, K., Pedersen, L.B., 1995. Collisional tectonics of the Baltic Shield in the northern Gulf of Bothnia from seismic data of the BABEL project. Geophys. J. Int. 120, 209–226.

Heikkinen, P., Luosto, U., 1992. Velocity structure and reflectivity of the Proterozoic crust in the Bothnian Sea. In: Meissner, R., Snyder, D., Balling, N., Staroste, E. (Eds.),

(47)

44

The BABEL Project, First Status Report, EUR 14429. Commission of the

European Communities, Directorate-General, Science, Research and Development, Brussels, pp. 65–69.

Högdahl, K., Sjöström, H., 2001. Evidence for 1.82 Ga transpressive shearing in a 1.85 Ga granitoid in central Sweden: implications for the regional evolution.

Precambrian Res. 105, 37–56.

Juhlin, C., Elming, S.-Å., Mellqvist, C., Öhlander, B., Weihed, P. & Wikström A. 2002.

Onshore crustal reflectivity of the Archaean-Proterozoic boundary and comparison with BABEL Lines 2 and 3, northern Sweden. Geophysical Journal International 150, 180–197.

Juhlin, C., Stephens, M., 2006. Gently dipping fracture zones in

Paleoproterozoic metagranite, Sweden: evidence from reflection seismic and cored borehole data, and implications for the disposal of nuclearwaste.

J.Geophys. Res.111, B09302. doi:10.1029/2005JB003887.

Ken Larnev, Ron Chambers, Mai Yang, Walt Lynn, and Willon Wai, 1983. Coherent noise in marine seismic data . GEOPHYSICS. 48, NO. 7.

Kohonen, J., 1995. From continental rifting to collisional shortening—Paleoproterozoic Kaleva metasediments of the Hoytiainen area in North Karelia, Finland. Geol. Surv.

Finland, Bull. 380, 79 pp.

Korhonen, J.V., Aaro, S., All, T., Nevanlinna, H., Skilbrei, J.R., Saavuori, H., Vaher, R., Zhdanova, L., Koistinen, T., 2002. Magnetic anomaly map of the Fennoscandian Shield 1:2 000 000. Geological Surveys of Finland, Norway and Sweden and Ministry of Natural Resources of Russian Federation.

Korja, A., Korja, T., Luosto, U. and Heikkinen, P., 1993. Seismic and geoelectric evidence for colisional and extensional events in the Fennoscandian Shield-implications

for Precambrian crustal evolution. Tectonophysics. 219, 129-152.

(48)

45

Korja, A., 1995. Structure of the Svecofennian crust—growth and destruction of the Svecofennian orogen. Ph.D. Thesis. Inst. Seismol., Univ. Helsinki, Re S-31, 36 p.

Korja, A., Heikkinen, P., Aaro, S., 2001. Crustal structure of the northern Baltic Sea paleorift.

Tectonophysics 331, 341– 358.

Korja, A. and Heikkinen, P., 2005. The accretionary Svecofennian orogen – insight from the BABEL profiles. Precambrian Research 136, 241-268.

Lahtinen, R., Korja, A., Nironen, M., Heikkinen, P., 2009. Palaeoproterozoic accretionary process in Fennoscandia. Geological Society, London, Special Publication v. 318; p.237-256 doi:10.1144/SP318.8.

Larner, K., Chambers, R., Yang, M., Lynn, W., Wai, W., 1983. Coherent noise in marine seismic data. Geophysics 48, 854-886.

Levin, F.K., 1971. Apparent velocity from dipping interface reflections: Geophysics, 36 , 510 – 516.

Malehmir, A., Dahlin, P., Lundberg, E., Juhlin, C., Sjöström, H., Högdahl, K., 2011.

Reflection seismic investigations in the Dannemora area, central Sweden: insights into the geometry of poly-phase deformation zones and magnetite-skarn deposits. Journal of Geophysical Research (JGR), vol. 116, B11307. doi:10.1029/2011JB008643.

Matthews, P.A., Graham, D.P., Long, R.E., 1992. Crustal structure beneath BABEL line 6 from wide-angle and normal-incidence seismic reflection data. In: Meissner, R., Snyder, D., Balling, N., Staroste, E. (Eds.), The BABEL Project, First Status Report,

EUR 14429. Commission of the European Communities, Directorate-General, Science, Research and Development, Brussels, pp. 101-104.

Mayne, W.H. 1962. Common reflection point horizontal data stacking technique.

Geophysics. 27:927-938.

Meissner, R., Wever, Th., Sadowiak, P., 1991. Continental collisions and seismic signature.

Geophys. J. Int. 105, 15–23.

(49)

46

Mooney, W.D., Brocher, T.M., 1987. Coincident seismic reflection/refraction studies of the continental lithosphere: a global review. Rev. Geophys. 25, 723–742.

Pedersen, L.B., Tryggvason, A., Schmidt, J., Gohl, K., 1992. Synthesis of geophysical data in the Bothnian Sea. In: Meissner, R., Snyder, D., Balling, N., Staroste, E. (Eds.), The BABEL Project, First Status Report, EUR 14429. Commission of the European Communitees, Directorate-General, Science, Research and Development, Brussels, pp. 59–64.

Rämö, T., Haapala, I., 1991. Tectonic setting and origin of the Proterozoic rapakivi granites of southeastern Fennoscandia. Royal Society of Edinburgh. 83, 165-171.

Sengbush, R.L., 1983. Seismic exploration methods: Internatinal Human Resources Development Corp., Boston.

Sheriff, R.E. and Geldart, L. P., 1995. Exploration seismology, Cambridge University Press.

Stork, C., Kapoor, J., Zhao, W., Dragoset, B., Dingwall, K., Western Geco, 2006. Predicting and removing complex 3D surface multiples with WEM modelling-an alternative to 3D SRME for wide azimuth surveys?

Society of Exploration Geophysicist, 2006-2679.

Yilmaz, Ö., 1989. Seismic data processing , Society of Exploration Geophysics, Tulsa.

Yilmaz, Ö., 2001. Seismic data analysis, Vol. 1, Society of Exploration Geophysics, Tulsa.

(50)

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

Time (s)

1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 11000 12000 13000 14000

W CDP E

20 Previous stack

TW T (S)

CDP Number

0 km 10

Appendix 1

(51)

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

Time (s)

1000 2000 3000 4000 5000 6000 7000

CDP

W E

Re-processed stack

TW T (S)

CDP Number

7000 E

20 0 km 10

Appendix 2

(52)

Tidigare utgivna publikationer i serien ISSN 1650-6553

Nr 1 Geomorphological mapping and hazard assessment of alpine areas in Vorarlberg, Austria, Marcus Gustavsson

Nr 2 Verification of the Turbulence Index used at SMHI, Stefan Bergman

Nr 3 Forecasting the next day’s maximum and minimum temperature in Vancouver, Canada by using artificial neural network models, Magnus Nilsson

Nr 4 The tectonic history of the Skyttorp-Vattholma fault zone, south-central Sweden, Anna Victoria Engström

Nr 5 Investigation on Surface energy fluxes and their relationship to synoptic weather patterns on Storglaciären, northern Sweden, Yvonne Kramer

Nr 230 Petrography and Geochemistry of a Section through the Blötberget Apatite-Iron Oxide Deposit, Bergslagen, South Central Sweden. Jiao Jingjing, December 2011

Nr 231 Structural Interpretation of Meyal Area (Upper Indus Basin), Pakistan Using Seismic Data. Abdullah Afzal Khan, Mars 2012

Nr 232 Analys av vinddata från lidar. Lisette Edvinsson, Mars 2012

Nr 233 Time Series Analysis of Ground Frost Conditions at Abisko, Sub- Arctic Sweden, 1985-2010, Anja Schmidt, May 2012

Nr 234 Temporal Dynamics of Total Organic Carbonexport Rates in Swedish Streams, Nino S Amvrosiadi, May 2012

Nr 235 Processing of Full Waveform Sonic Data for Shear Wave Velocity at the Ketzin CO² Storage Site, Khalid Abbas, May 2012

Nr 236 Transport of Multi-Walled Carbon Nanotubes in Saturated Porous Media Dixiao Bao, June 2012

Nr 237 Structural Model of the Lambarfjärden Area from Surface and Subsurface Data in Connection with the E4 Stockholm Bypass Anna Vass, June 2012

Nr 238 Mechanisms Controlling Valley Asymmetry Development at Abisko, Northern Sweden and Sani Pass, Southern Africa, Carl-Johan Borg, August 2012

Nr 239 Effect of Orientation on Propagation of Pre-existing fractures, Hajab Zahra, August 2012

Nr 240 Mobility of multi-walled carbon nanotubes in unsaturated porous media, Abenezer Mekonen, August 2012

(53)

Re-processing of Shallow and Deep Crustal Reflection Seismic Data along BABEL Line 7, Central Sweden

Examensarbete vid Institutionen för geovetenskaper ISSN 1650-6553 Nr 241